Japanese Anime Speech: 10K-100K Audio-Text Pairs from Visual Novels

Japanese Anime Speech contains between 10,000 and 100,000 audio-text pairs sourced from Japanese visual novels, released by joujiboi in 2024. The collection pairs audio clips with their corresponding transcriptions to improve the accuracy of automatic speech recognition models for anime-style media.

Use Cases

• Fine-tuning OpenAI Whisper models using the audio clips and transcriptions
• Training Japanese automatic speech recognition (ASR) models for media-specific dialogue
• Benchmarking speech-to-text performance on visual novel audio

Strengths

• Contains 10,000 to 100,000 audio-text pairs
• CC0 1.0 public domain license
• Domain-specific focus on visual novel audio

Limitations

• Stylized visual novel dialogue may not reflect natural Japanese speech patterns
• Lack of metadata regarding speaker age, gender, or recording environment

Provenance

Source

Visual novels

Collection Method

Extracted from visual novel media

Freshness

Last updated September 2024.

Geography

Japan